52 lines
1.6 KiB
Python
52 lines
1.6 KiB
Python
|
# Adapted from:
|
||
|
|
# Copyright (c) 2014 Rowan Nairn. All rights reserved. Licensed under the 2-clause BSD licence
|
||
|
|
# https://github.com/r0wb0t/markdown-urlize
|
||
|
|
|
||
|
|
import markdown
|
||
|
|
|
||
|
|
# Global Vars
|
||
|
|
URLIZE_RE = '(%s)' % '|'.join([
|
||
|
|
r'<(?:f|ht)tps?://[^>]*>',
|
||
|
|
r'\b(?:f|ht)tps?://[^)<>\s]+[^.,)<>\s]',
|
||
|
|
r'\bwww\.[^)<>\s]+[^.,)<>\s]',
|
||
|
|
r'[^(<\s]+\.(?:(?:com|net|org)(?:\.au)?)\b',
|
||
|
|
r'doi: *(?:[^ ]+)'
|
||
|
|
])
|
||
|
|
|
||
|
|
class UrlizePattern(markdown.inlinepatterns.Pattern):
|
||
|
|
""" Return a link Element given an autolink (`http://example/com`). """
|
||
|
|
def handleMatch(self, m):
|
||
|
|
url = m.group(2)
|
||
|
|
|
||
|
|
if url.startswith('<'):
|
||
|
|
url = url[1:-1]
|
||
|
|
|
||
|
|
text = url
|
||
|
|
|
||
|
|
if not url.split('://')[0] in ('http','https','ftp'):
|
||
|
|
if '@' in url and not '/' in url:
|
||
|
|
url = 'mailto:' + url
|
||
|
|
elif url.startswith('doi:'):
|
||
|
|
url = 'https://doi-org.ezproxy.lib.monash.edu.au/' + url[4:].strip()
|
||
|
|
else:
|
||
|
|
url = 'http://' + url
|
||
|
|
|
||
|
|
el = markdown.util.etree.Element("a")
|
||
|
|
el.set('href', url)
|
||
|
|
el.text = markdown.util.AtomicString(text)
|
||
|
|
return el
|
||
|
|
|
||
|
|
class UrlizeExtension(markdown.Extension):
|
||
|
|
""" Urlize Extension for Python-Markdown. """
|
||
|
|
|
||
|
|
def extendMarkdown(self, md, md_globals):
|
||
|
|
""" Replace autolink with UrlizePattern """
|
||
|
|
md.inlinePatterns['autolink'] = UrlizePattern(URLIZE_RE, md)
|
||
|
|
|
||
|
|
def makeExtension(*args, **kwargs):
|
||
|
|
return UrlizeExtension(*args, **kwargs)
|
||
|
|
|
||
|
|
if __name__ == "__main__":
|
||
|
|
import doctest
|
||
|
|
doctest.testmod()
|